Definitive MPLS Network Designs [Electronic resources] نسخه متنی

اینجــــا یک کتابخانه دیجیتالی است

با بیش از 100000 منبع الکترونیکی رایگان به زبان فارسی ، عربی و انگلیسی

Definitive MPLS Network Designs [Electronic resources] - نسخه متنی

Jim Guichard; François Le Faucheur; Jean-Philippe Vasseur

| نمايش فراداده ، افزودن یک نقد و بررسی
افزودن به کتابخانه شخصی
ارسال به دوستان
جستجو در متن کتاب
بیشتر
تنظیمات قلم

فونت

اندازه قلم

+ - پیش فرض

حالت نمایش

روز نیمروز شب
جستجو در لغت نامه
بیشتر
لیست موضوعات
توضیحات
افزودن یادداشت جدید





Carrier's Carrier Architecture


The initial Layer 3 MPLS VPN service consisted of the architecture detailed in the previous sections and was primarily targeted at providing Layer 3 network-based VPN services to customers with a limited amount of IP routing information (because of their relative size in terms of regional and global reach). With restricted routing information, the service provider PE routers could scale to a large number of end customers. This model is still the primary delivery vehicle for Layer 3 MPLS VPN services today.

However, over time larger customer deployments have been requested, including large Enterprises and Internet service providers (ISPs), as well as other VPN providers, that want to obtain IP-based any-to-any connectivity. This requirement has been driven in part by the desire of these customers not to have to build their own Layer 2 network, but instead to use the Layer 3 MPLS VPN network for transport. With the reduction in infrastructure costs and the added benefit of full any-to-any connectivity, this service is very attractive.

Although the service is attractive for large Enterprises and other carriers for the reasons already noted, it comes at a considerable cost, in terms of scale, to the operator of the Layer 3 MPLS VPN network. For example, consider an ISP as the customer of this service. It might want to exchange full Internet routes (some 155,000+ routes) between its sites. Because the Layer 3 MPLS VPN service requires that customer routing state be stored at the PE routers and be distributed to other PE routers within its control, this presents a significant scaling problem at the edge of the network, not to mention the route distribution load offered to the control plane within the service provider network.

As a trade-off to provide connectivity for customers with large routing requirements, and as a complement to the regular Layer 3 MPLS VPN services for other VPN or non-VPN carriers that want to obtain basic IP connectivity, the Carrier's Carrier architecture was introduced. This architecture allows for the direct exchange of customer routes between their own sites and the use of the Layer 3 MPLS VPN backbone as pure IP "transport" between these sites. Figure 1-12 illustrates this concept.


Figure 1-12. Carrier's Carrier Architecture

Chapter 4, "National Telco Design Study."

In [BGP+Label]).

Carrier network The network of the carrier that is obtaining VPN services from the MPLS VPN Service Provider. This network may or may not be enabled to run MPLS switching functionality, either with LDP, TDP, or RSVP-TE.

Carrier's Carrier network The Layer 3 MPLS VPN network that provides Carrier's Carrier functionality at its PE routers for attached Carrier networks.


Because all external routing information is exchanged directly between the customer sites, there is no need for the service provider that provides Carrier's Carrier services to carry the routes at its PE routers. If you consider the ISP example of 155,000+ Kbps routes, removing the requirement to carry these routes substantially reduces the amount of routing state held by the PE routers. However, because the BGP-4 next-hop addresses for these routes will be within the customer routing space, these addresses need to be exchanged with the Carrier's Carrier service provider so that an end-to-end packet-forwarding path can be achieved.


Packet Forwarding with Carrier's Carrier


Because only internal routes are exchanged between the CSC CE routers of the carrier network and the CSC PE routers of the Carrier's Carrier network, the CSC PE routers no longer carry any customer external routing information and therefore can't forward IP packets toward those destinations. The result is that the CSC PE routers must be able to forward packets using a label [BGP+Label]) between the CSC CE routers and CSC PE routers. All these elements are shown in Figure 1-13.


Figure 1-13. Internal/External Route Exchange

[View full size image]

Having exchanged all the internal and external routing information, an end-to-end LSP may be established between the CSC CE routers. Figure 1-14 shows the LSP creation between two customer sites for traffic sent toward subnet 196.1.1.0/24. OL refers to the outer label, and VL refers the VPN inner label.


Figure 1-14. LSP Between CSC Customer Sites

Chapter 4. This does not affect the overall packet forwarding scheme details in this section. It simply means that the LSP used to establish an end-to-end LSP between carrier sites is extended beyond the CSC CE routers.

A more detailed analysis of the Carrier's Carrier architecture may be found in [VPN-ARCH-Volume-2].


/ 96